I
would be a little more thrilled if this were to be slightly
predictive. More likely it is outright misleading, but it does
remind us that the whole pattern of macroscopic variation is a
primary third order relationship which is really a binary
relationship held in place by the strength of the PNA pattern.
In
the meantime delta heat has been injected on effectively a constant
basis since the late seventies at least if not since the early
sixties. This is easiest accomplished by the agency of the sea for
which we even have conforming evidence. That at least allowed me to
predict a primary collapse of sea ice by 2012 as early as 2007.
I
am more and more convinced that the atmospheric aspect of arctic
warming is secondary to the impact of ocean currents that also vary
but over vastly longer time cycles. Recall that shutting off the gulf
stream would encase the north Polar Ice Cap in a raging Ice Age at
least to fifteen degrees south and that ice floes would flood Hudson
Bay all year round.
Saturday,
June 22, 2013
A
new paper published in the International
Journal of Climatology finds
40-79% of the variation in western Arctic sea ice extent is due to
natural atmospheric circulation patterns such as the Arctic Dipole
(AD), Arctic Oscillation (AO), and Pacific-North American (PNA)
pattern.
Prior research has shown Arctic sea ice extent is also controlled by
natural variation of winds
and
storm
activity,
and related to the geography
of
the Arctic, not alleged anthropogenic global warming.
Thomas J. Ballinger, Scott C. Sheridan
ABSTRACT: In
this study, a synoptic climatological approach is employed to assess
the relationship between the frequency of circulation patterns (CPs)
and the latitude of mid-September sea ice minima in the western
Arctic. Fifteen CPs are created via principal component analysis and
cluster analysis from daily NCEP/NCAR reanalysis sea-level pressure
(SLP) fields across a grid from 50 to 90°N and 150°E–100°W from
1979 to 2011. The frequency of these CPs are statistically compared
with the latitude of the sea ice minimum from passive microwave data
for each of 11 equally-spaced longitudes (176°W to 126°W) extending
into the Chukchi and Beaufort Seas. Monthly frequencies for each of
the 15 CPs from March to September, signifying the melt season, for
each year are correlated with the ice minimum for that September.
These monthly frequencies are then entered into a stepwise multiple
linear regression (SMLR) and collectively, CP
[circulation pattern] frequencies explain 40–79% of the total ice
retreat variance across the longitudes. The
frequency of one cluster, CP 11, representing a broad high pressure
area over the Beaufort Sea, is highly correlated with the latitude of
the sea ice minima; June and August frequencies of this pattern are
the initial predictors at 8 of the 11 longitudes and explain 22–32%
of the variance. This pattern has occurred more frequently from 2007
onwards; compared with a June mean occurrence of 9 days during
1979–2006, CP 11 occurred 16 times in June 2007, and on average
more than 17 days per month during June 2008–2011. The
Arctic Dipole (AD), Arctic Oscillation (AO), and Pacific-North
American (PNA) pattern indices are significantly correlated with CPs
11–13 frequencies throughout certain summer months, further
indicating strong relationships between summer circulation and sea
ice minima in the region.
I stopped reading when I saw the graph. There are liars, damn liars, and statisticians. Figures do not lie, but liars figure. There are a lot of ways to lie, and one of the most insidious is by omission.
ReplyDeleteYou omitted the two standard deviation block. That renders the entire graph a lie.
Go here to see something more accurate
http://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/